Module 06 · Tier 4 — Intermediate Prompting

Writing with AI at Pro Level

From first draft to final polish. How to keep your voice while multiplying your output.

12 min Tier 4 — Intermediate Prompting

Why This Matters

Pro-level AI writing isn't about volume — it's about quality and authenticity at scale. The writers who use AI most effectively produce work that is indistinguishable from their best unassisted output, in a fraction of the time. The ones who produce obvious AI slop have outsourced the thinking and kept only the typing. This module is about the difference — how to use AI as a genuine writing multiplier without losing what makes your writing yours.

AI's default writing voice is optimized for plausibility, not distinctiveness. It's articulate, balanced, and forgettable. It uses the vocabulary of competent corporate communication: "leverage," "implement," "ensure alignment," "stakeholders." It qualifies everything. It never takes a strong position without immediately hedging it.

Your voice is the opposite of this — it has opinions, preferences, rhythms, and tells. Preserving it while using AI requires deliberate technique, not hope.

Before asking AI to write anything in your voice, paste 3-5 examples of your own writing and say: "Match the tone, rhythm, and vocabulary of these examples. Note what makes them distinctive and apply that to your response." This works better than any description you could give.

Full access is for AIQ members

Unlock all 56 lessons, the certificate pathway, and the SociA|~ community.

56 lessons across all three course tracks
AIQ certification on completion
SociA|~ Society community access

Unlock full access →

The Concept

The Voice Preservation Problem

Your voice is the opposite of this — it has opinions, preferences, rhythms, and tells. Preserving it while using AI requires deliberate technique, not hope.

The Four-Layer Voice System

Layer 1: Sample injection

Layer 2: Anti-vocabulary lists

Every writer has words and phrases they never use. AI has default words it always uses. Build a short list of AI defaults to prohibit: "Do not use: leverage, synergy, implement, ensure, stakeholders, cutting-edge, game-changing, innovative." This alone dramatically reduces generic-sounding output.

Layer 3: Position clarity

AI defaults to "on one hand / on the other hand" balance. If your voice takes positions, tell it to: "Take a clear position and defend it. Don't hedge unless the evidence genuinely warrants uncertainty." If your voice is genuinely balanced, tell it that explicitly too.

Layer 4: Edit as the final step

No matter how well-calibrated your prompt, the last 20% of voice recovery happens in editing. Treat AI's output as a strong first draft, not a finished piece. Read it aloud — where does it sound like you? Where does it sound like a committee wrote it? Fix the second category.

The Output Multiplier Framework

Pro AI writers think in terms of one piece of core thinking generating multiple outputs:

Write the core argument / key insight once, in your own words
Ask AI to adapt it: "Turn this into a LinkedIn post," "make this a 3-slide executive summary," "write this as an email to a skeptical CFO"
Edit each adaptation for voice and accuracy

One strong original insight → 4-6 adapted outputs, each edited → a week of content produced in a morning. This is the actual AI writing multiplier, not "have AI write it from scratch."

The Two Things AI Does Better Than You

Even excellent writers should use AI for two specific writing tasks where AI consistently outperforms human first-pass attempts:

Structural reordering: "Here is my draft. Suggest three different ways to reorder the argument. For each one, explain what it emphasizes differently." Humans are attached to the order they wrote in. AI isn't.

Audience translation: "Rewrite this for an audience of [specific audience]. Assume they care about [specific concern] and are skeptical about [specific thing]." This takes judgment to do well — AI can do a credible version in seconds.

Voice preserved vs. voice lost

Same prompt, two approaches:

Voice-lost approach: "Write a LinkedIn post about the importance of AI fluency in the workplace."

Result: "In today's rapidly evolving digital landscape, AI fluency has become an essential skill for modern professionals. Organizations that embrace AI-driven transformation are better positioned to leverage competitive advantages..." — forgettable, generic, indistinguishable from ten thousand other posts.

Voice-preserved approach: "Here are three of my previous LinkedIn posts: [paste examples]. Match this voice exactly. Write a post about AI fluency at work. Take a specific, opinionated position rather than hedging. Start with a specific observation, not a general statement. Under 200 words. Do not use: leverage, evolving landscape, digital transformation, essential skills."

Result: Something that sounds like a specific person with a specific perspective, starting with a concrete observation, making a real argument. Shareable. Memorable.

Hands-On Exercise

Build your personal voice calibration prompt

ClaudeChatGPT

This exercise produces something you'll use permanently: a voice calibration prompt you can apply to any AI writing task. Step 1: Find 3-4 pieces of your own writing that you're proud of — emails, documents, posts, anything. Step 2: Ask Claude or ChatGPT: "Read these examples of my writing. What are five words or phrases I use frequently? What is the rhythm of my sentences (long/short, varied/consistent)? What position do I take — direct or hedged? What vocabulary do I avoid?" Step 3: Read the AI's analysis. What did it get right? Edit it until it's accurate. Step 4: Build your calibration prompt: "Write in [your name]'s voice. Characteristics: [paste the accurate description]. Never use: [your anti-vocabulary list]." Test it on a real writing task. Compare to your unassisted writing. Edit until the gap is small.

The calibration prompt is worth 30 minutes to build properly. You'll use it for years.

Active Recall

Before moving on — close this lesson and answer these from memory. Then come back and check. Testing yourself (not re-reading) is how this sticks.

1 What are the four layers of the voice preservation system? Why is editing as the final step non-optional?

2 What is the output multiplier framework? How does it differ from asking AI to write content from scratch?

Reflection

What is the gap between your best unassisted writing and your typical AI-assisted writing? Which of the four voice preservation layers would close that gap the most?

Key Takeaway

Pro AI writing preserves your voice through sample injection, anti-vocabulary lists, position clarity, and aggressive editing. The multiplier framework: one strong original insight, AI-adapted into multiple formats, each edited. Core thinking is always yours.